test-float/test-ifloat failure in yn

[Roland McGrath]

On my machine (Pentium III), there is this error in the test suite.
Is this a known problem?

Failure: Test: yn (10, 2.0) == -129184.542208039282635913145923304214
Result:
 is:         -1.29184562500000000000e+05  -0x1.f8a09000000000000000p+16
 should be:  -1.29184539062500000000e+05  -0x1.f8a08a00000000000000p+16
 difference:  2.34375000000000000000e-02   0x1.80000000000000000000p-6
 ulp       :  3.0000
 max.ulp   :  1.0000